Some Begram Ivories and the South Indian Narrative Tradition: New Evidence Elizabeth Rosen Stone Many of the Begram ivories rank amongst the finest works lication of the hoard in the Mémoires de la Délégation of art surviving from ancient India. Despite their profound archéologiques française en Afghanistan 9 appeared shortly interest, they are relatively small pieces of ivory and bone made thereafter.3 With regard to the ivories, Hackin compared some in a wide variety of styles and techniques and probably by a of them to Stupa I at Sanchi, while one very important one, large number of craftsmen from different regions and even at Coffret IX (Fig. 6),4 he placed from the end of the third to the different times. The ivories were used to decorate wooden fur- beginning of the fourth century CE.5 His second publication niture, the wood having long since disintegrated. Many of was posthumous, as he and his wife were both killed in 1941, them were imported from India and may have arrived in but their notes were left in safe hands. By 1954, the second Begram already assembled into furniture while others may publication appeared in volume 11 of the same series with have been assembled somewhere near Begram, in eastern Bac- important articles by Philippe Stern, Jeannine Auboyer, Otto tria near the border of Gandhara (Map 1). The ivories were Kurz, and others, as well as important drawings by Pierre found in two sealed rooms, numbers 10 and 13, along with Hamelin. 6 Emphasizing a different group of ivories, and using numerous other treasures that included bronzes and plaster different comparative material, Philippe Stern assigned most casts of metalwork imported from the Greco-Roman world, of the ivories to the first and second centuries CE but admitted Alexandrian glass, and lacquer work from China.1 Neither the that Coffret IX appeared stylistically somewhat later.7 These site of the Begram hoard nor the accumulation of objects con- two studies, as well as the excavation notes, remain the founda- tained within it provides us with definitive dates. The hoard is tion upon which all subsequent studies have been made. That culturally heterogeneous which certainly suggests that it is the Hackins were killed in World War II was not only a human chronologically heterogeneous. tragedy but also an archaeological one, for it seems that there The style of many of the ivories is Indian. Over the years, was much to learn from further excavation. In the 1970s, a there has been much controversy as to their place of manufac- team from the Archaeological Survey of India planned re- ture, as well as their date. They have been assigned to dates excavate Begram, but unfortunately those plans had to be ranging from the first century BCE to the fourth century CE abandoned.8 and have been compared to almost every major site of sculp- The majority of the Begram treasure was housed in the tural production within the Indian subcontinent. The reasons Kabul Museum, and many of its objects were lost or damaged for this are intrinsic to the field of Indian art and archaeology. during the war and civil unrest in Afghanistan. Despite this, Since only a small percentage of ancient sites have been identi- new studies have made the ivories easier to study, at least from fied, excavated, and published, we are constantly restudying photographs, than ever before. In the recently published cata- the same evidence over and over in hopes of somehow “re- logue of the National Museum of Afghanistan,9 covering the dating” or “reassessing” them. Furthermore, the major regional objects there from 1935 through 1985, Francine Tissot presents sculptural schools were connected by trade routes and influ- the ivories along with the proposed reconstructions of the fur- enced each other so that styles frequently tended to comingle. niture in which they were set, helping us to understand their Thus, despite the existence of clear regional styles, we often see context. Wisely, she cautions us when the reconstructions are forms or motifs in different regions at the same time.2 Further speculative. The catalogue, complete with inventory numbers complicating the problem, India’s reverence for the past makes and references, also includes some unpublished material. it difficult to separate that which is truly old from that which While she dates the ivories to the first and second centuries or is merely archaizing. To date, none of the theories regarding perhaps a little later,10 the emphasis of the book is documenta- the ivories are universally accepted, and I do not propose to tion rather than a discussion of the dating. definitively solve the problems of date and/or provenance in The detailed doctoral dissertation of Sanjyot Mehendale,11 this paper. What I will do is to insist upon the importance of which is available on a website,12 continues the process of the southern Indian contribution to the style and motifs of the organization. She separates the ivories into categories of sub- Begram ivories ject matter and technique and brings together a vast amount The Begram hoard was first excavated between 1937 and of interpretive information. While she draws her own conclu- 1939 by Joseph Hackin and his wife Ria, and their initial pub- sions, she generously cites those taking other positions, creat- jiaaa3n.indd 45 13-10-2009 08:04:31 ing an invaluable resource. Based upon recent archaeological its style were probably transmitted by itinerant artists to many evidence she has introduced the notion of a Central Asian other sites.21 One such site, Kanganhalli in Karnataka (Map component to several of the ivories, a subject that was not part 2), has only recently begun to be known. It has been excavated of the earlier literature or archaeology. Mehendale concluded by the Archaeological Survey of India, and we are looking for- that the ivories were produced in the first century CE in the ward to its publication. Nevertheless, I wish to present a por- northwest, perhaps even in Begram itself, by artists trained in tion of this material at the present time. While we must await or conversant with other traditions.13 Her study provoked a further studies of the Kanganhalli material for secure dates, strong response from Lolita Nehru.14 the visual comparisons with the Begram ivories are striking and Using other evidence, Nehru placed the origin of some of further reinforce the hypothesis that a group of the ivories had the ivories in Mathura in the second century CE. She also con- their stylistic sources in southern India.22 tinued to address and expand upon the question of a Central Kanganhalli, first excavated in 1994, is a highly significant Asian affiliation and reverts somewhat to Hackin’s original stupa site of which a large portion of the sculptural elements dating, bringing it up to the early third century CE.15 For one survive. The sculptures provide us with everything we have familiar with the art and issues of the Begram ivories, it is long been looking for: sources and comparative material for tempting to address each point individually; but that is not my the Andhra sculptural tradition, as well as for some of the intention. The avowed purpose of this article is simply to add Begram ivories. It is only a few kilometers from the more well- some new information in a timely manner- and yet another known site of Sannathi23 on the east bank of the Bhima River, article!- to the already vast bibliography in hopes that one day just south of Gulbarga. Thus both Sannathi and Kanganhalli we will have a better understanding of the many complex issues were farthest west of the Buddhist establishments along a vast surrounding the ivories. river system ending in the east coast of Andhra Pradesh The My own interest in the Begram ivories began in the 1970s two sites should be considered together, as well as with the while I was researching my doctoral dissertation on the Andhra sites. Like most Indian stupas, the one at Kanganhalli Buddhist narrative art of Nagarjunakonda, produced in An- was built over a long period of time with continuous additions dhra Pradesh (Map 2) between c. 225 and c. 325 CE. The clear of both narrative reliefs and free-standing sculptures. The most evidence of Roman trade and influence are seen in these Bud- important available information about Kanganhalli appears dhist narratives, and perhaps there were even Roman crafts- on the internet in two descriptive articles by D. Dayalan of the men at Nagarjunakonda as reflected in the architecture of the Archaeological Survey of India.24 According to him, the drum so-called stadium.16 The style of Nagarjunakonda was previ- of the stupa was in two tiers. While I am unable to judge defin- ously considered a unique outgrowth of the school of Amara- itively from the photographic documentation alone, it would vati, but recent excavations at the site of Phanigiri 17 show that seem that the stupa could resemble the one excavated at Chan- the Nagarjunakonda idiom was more widely spread than hith- davaram in the Prakasam District of Andhra Pradesh.25 In the erto known. The many parallels between the Begram ivories Kanganhalli stupa, the lower portion is decorated with archi- and the art of Nagarjunakonda - Indian style and foreign in- tectural representations alternating with blank slabs, while the terpolations – brought me to the Kabul Museum to study decoration of the upper portion contains tall narrow slabs with these ivories.18 Shortly thereafter, I published the results of my jatakas and scenes from the life of the Buddha. Panels of a research in a student journal, Marsyas, in which I proposed a similar shape were excavated at Chandavaram.